What problem does it solve? Teams building event-driven services on Kafka often ship inconsistent topic names, unreliable producers, unsafe consumer offset handling, and breaking schema changes that cause silent message loss or consumer failures. This Skill provides a concrete rule set that prevents those failure modes before code is written or reviewed. ## Core Features & Use Cases - Topic and Schema Conventions: Enforces domain.context.eventName.v1 topic naming and Schema Registry compatibility (BACKWARD/BACKWARD_TRANSITIVE) with safe schema evolution rules. - Producer and Consumer Reliability Rules: Mandates idempotent producers (enable.idempotence=true, acks=all), keyed records for ordering, manual offset commits, and idempotent consumer processing. - DLQ and Poison-Pill Handling: Requires every consumer to route unprocessable messages to a dead-letter topic instead of retrying forever, with Java/Spring Kafka code examples. - Use Case: When implementing a new Kafka consumer for order events, apply these rules to configure manual acknowledgment, route poison-pill messages to the DLQ, and propagate correlation IDs via message headers. ## Quick Start Apply the messaging-kafka rules to review my Kafka producer and consumer configuration for reliability and schema compatibility issues.
